Machine learning at the mesoscale: A computation-dissipation bottleneck.

The cost of information processing in physical systems calls for a trade-off between performance and energetic expenditure. Here we formulate and study a computation-dissipation bottleneck in mesoscopic systems used as input-output devices. Using both real data sets and synthetic tasks, we show how nonequilibrium leads to enhanced performance. Our framework sheds light on a crucial compromise between information compression, input-output computation and dynamic irreversibility induced by nonreciprocal interactions.

Collective foraging of active particles trained by reinforcement learning.

Collective self-organization of animal groups is a recurring phenomenon in nature which has attracted a lot of attention in natural and social sciences. To understand how collective motion can be achieved without the presence of an external control, social interactions have been considered which regulate the motion and orientation of neighbors relative to each other. Here, we want to understand the motivation and possible reasons behind the emergence of such interaction rules using an experimental model system of light-responsive active colloidal particles (APs). Via reinforcement learning (RL), the motion of particles is optimized regarding their foraging behavior in presence of randomly appearing food sources. Although RL maximizes the rewards of single APs, we observe the emergence of collective behaviors within the particle group. The advantage of such collective strategy in context of foraging is to compensate lack of local information which strongly increases the robustness of the resulting policy. Our results demonstrate that collective behavior may not only result on the optimization of behaviors on the group level but may also arise from maximizing the benefit of individuals. Apart from a better understanding of collective behaviors in natural systems, these results may also be useful in context of the design of autonomous robotic systems.

Seeking and sharing information in collective olfactory search.

Searching for a target is a task of fundamental importance for many living organisms. Long-distance search guided by olfactory cues is a prototypical example of such a process. The searcher receives signals that are sparse and very noisy, making the task extremely difficult. Information-seeking strategies have thus been proven to be effective for individual olfactory search and their extension to collective search has been the subject of some exploratory studies. Here, we study in detail how sharing information among members of a group affects the search behavior when agents adopt information-seeking strategies as Infotaxis and its recently introduced variant, Space-Aware Infotaxis. Our results show that even in absence of explicit coordination, sharing information leads to an effective partitioning of the search space among agents that results in a significant decrease of mean search times.

Olfactory search with finite-state controllers.

Long-range olfactory search is an extremely difficult task in view of the sparsity of odor signals that are available to the searcher and the complex encoding of the information about the source location. Current algorithmic approaches typically require a continuous memory space, sometimes of large dimensionality, which may hamper their optimization and often obscure their interpretation. Here, we show how finite-state controllers with a small set of discrete memory states are expressive enough to display rich, time-extended behavioral modules that resemble the ones observed in living organisms. Finite-state controllers optimized for olfactory search have an immediate interpretation in terms of approximate clocks and coarse-grained spatial maps, suggesting connections with neural models of search behavior.

Stable cooperation emerges in stochastic multiplicative growth.

Understanding the evolutionary stability of cooperation is a central problem in biology, sociology, and economics. There exist only a few known mechanisms that guarantee the existence of cooperation and its robustness to cheating. Here, we introduce a mechanism for the emergence of cooperation in the presence of fluctuations. We consider agents whose wealth changes stochastically in a multiplicative fashion. Each agent can share part of her wealth as a public good, which is equally distributed among all the agents. We show that, when agents operate with long-time horizons, cooperation produces an advantage at the individual level, as it effectively screens agents from the deleterious effect of environmental fluctuations.

Dynamics and risk sharing in groups of selfish individuals.

Understanding why animals organize in collective states is a central question of current research in, e.g., biology, physics, and psychology. More than 50 years ago, W.D. Hamilton postulated that the formation of animal herds may simply result from the individual's selfish motivation to minimize their predation risk. The latter is quantified by the domain of danger (DOD) which is given by the Voronoi area around each individual. In fact, simulations show that individuals aiming to reduce their DODs form compact groups similar to what is observed in many living systems. However, despite the apparent simplicity of this problem, it is not clear what motional strategy is required to find an optimal solution. Here, we use the framework of Multi Agent Reinforcement Learning (MARL) which gives the unbiased and optimal strategy of individuals to solve the selfish herd problem. We demonstrate that the motivation of individuals to reduce their predation risk naturally leads to pronounced collective behaviors including the formation of cohesive swirls. We reveal a previously unexplored rather complex intra-group motion which eventually leads to a evenly shared predation risk amongst selfish individuals.

Frictionless nanohighways on crystalline surfaces.

The understanding of friction at nano-scales, ruled by the regular arrangement of atoms, is surprisingly incomplete. Here we provide a unified understanding by studying the interlocking potential energy of two infinite contacting surfaces with arbitrary lattice symmetries, and extending it to finite contacts. We categorize, based purely on geometrical features, all possible contacts into three different types: a structurally lubric contact where the monolayer can move isotropically without friction, a corrugated and strongly interlocked contact, and a newly discovered directionally structurally lubric contact where the layer can move frictionlessly along one specific direction and retains finite friction along all other directions. This novel category is energetically stable against rotational perturbations and provides extreme friction anisotropy. The finite-size analysis shows that our categorization applies to a wide range of technologically relevant materials in contact, from adsorbates on crystal surfaces to layered two-dimensional materials and colloidal monolayers.

Pervasive orientational and directional locking at geometrically heterogeneous sliding interfaces.

Understanding the drift motion and dynamical locking of crystalline clusters on patterned substrates is important for the diffusion and manipulation of nano- and microscale objects on surfaces. In a previous work, we studied the orientational and directional locking of colloidal two-dimensional clusters with triangular structure driven across a triangular substrate lattice. Here we show with experiments and simulations that such locking features arise for clusters with arbitrary lattice structure sliding across arbitrary regular substrates. Similar to triangular-triangular contacts, orientational and directional locking are strongly correlated via the real- and reciprocal-space Moiré patterns of the contacting surfaces. Due to the different symmetries of the surfaces in contact, however, the relation between the locking orientation and the locking direction becomes more complicated compared to interfaces composed of identical lattice symmetries. We provide a generalized formalism which describes the relation between the locking orientation and locking direction with arbitrary lattice symmetries.

Pile-up transmission and reflection of topological defects at grain boundaries in colloidal crystals.

Crystalline solids typically contain large amounts of defects such as dislocations and interstitials. How they travel across grain boundaries (GBs) under external stress is crucial to understand the mechanical properties of polycrystalline materials. Here, we experimentally and theoretically investigate with single-particle resolution how the atomic structure of GBs affects the dynamics of interstitial defects driven across monolayer colloidal polycrystals. Owing to the complex inherent GB structure, we observe a rich dynamical behavior of defects near GBs. Below a critical driving force defects cannot cross GBs, resulting in their accumulation near these locations. Under certain conditions, defects are reflected at GBs, leading to their enrichment at specific regions within polycrystals. The channeling of defects within samples of specifically-designed GB structures opens up the possibility to design novel materials that are able to confine the spread of damage to certain regions.

Velocity dependence of sliding friction on a crystalline surface.

We introduce and study a minimal 1D model for the simulation of dynamic friction and dissipation at the atomic scale. This model consists of a point mass (slider) that moves over and interacts weakly with a linear chain of particles interconnected by springs, representing a crystalline substrate. This interaction converts a part of the kinetic energy of the slider into phonon waves in the substrate. As a result, the slider experiences a friction force. As a function of the slider speed, we observe dissipation peaks at specific values of the slider speed, whose nature we understand by means of a Fourier analysis of the excited phonon modes. By relating the phonon phase velocities with the slider velocity, we obtain an equation whose solutions predict which phonons are being excited by the slider moving at a given speed.

Ballistic thermophoresis of adsorbates on free-standing graphene.

The textbook thermophoretic force which acts on a body in a fluid is proportional to the local temperature gradient. The same is expected to hold for the macroscopic drift behavior of a diffusive cluster or molecule physisorbed on a solid surface. The question we explore here is whether that is still valid on a 2D membrane such as graphene at short sheet length. By means of a nonequilibrium molecular dynamics study of a test system-a gold nanocluster adsorbed on free-standing graphene clamped between two temperatures [Formula: see text] apart-we find a phoretic force which for submicron sheet lengths is parallel to, but basically independent of, the local gradient magnitude. This identifies a thermophoretic regime that is ballistic rather than diffusive, persisting up to and beyond a 100-nanometer sheet length. Analysis shows that the phoretic force is due to the flexural phonons, whose flow is known to be ballistic and distance-independent up to relatively long mean-free paths. However, ordinary harmonic phonons should only carry crystal momentum and, while impinging on the cluster, should not be able to impress real momentum. We show that graphene and other membrane-like monolayers support a specific anharmonic connection between the flexural corrugation and longitudinal phonons whose fast escape leaves behind a 2D-projected mass density increase endowing the flexural phonons, as they move with their group velocity, with real momentum, part of which is transmitted to the adsorbate through scattering. The resulting distance-independent ballistic thermophoretic force is not unlikely to possess practical applications.

Nanoscale Effects on Phase Separation.

Classical nucleation theory predicts that a binary system which is immiscible in the bulk should become miscible at the nanoscale when lowering its size below a critical size. Here we tackle the problem of miscibility in nanoalloys with a combination of ab initio and atomistic calculations, developing a statistical-mechanics approach for the free energy cost of forming phase-separated aggregates. We apply it to the controversial case of AuCo nanoalloys. AuCo is immiscible in the bulk, but a rich variety of nanoparticle configurations, both phase-separated and intermixed, have been obtained experimentally. Our calculations strongly point to the permanence of an equilibrium miscibility gap down to the nanoscale and to the nonexistence of a critical size below which phase separation is impossible. We show that this is due to nanoscale effects of general character, caused by the existence of preferred nucleation sites in nanoparticles, which lower the free-energy cost for phase separation with respect to bulk systems.

Strain-induced restructuring of the surface in core@shell nanoalloys.

A general trend towards surface reconstructions which eliminate open facets in favour of close-packed facets is found for a wide class of A@B nanoalloys with atomically thin shells (B-skin structures), by means of a combination of atomistic and ab initio density-functional calculations. This class comprises Co@Au, Co@Pt, Ni@Pt, Co@Ag, Ni@Ag, Cu@Ag, Cu@Au, Rh@Au, Ni@Rh and Ni@Pd nanoalloys, in which core atoms are smaller than shell atoms. The reconstruction can induce a global restructuring of the nanoparticles, with different parts transforming in a concerted way and causing the emergence of pyritohedral structures from fcc nanoalloys, and of chiral decahedra. The instability of open facets is rationalized in terms of the equilibration of atomic pressures in strained structures. The surface reconstruction can have important effects on nanoalloy properties, as shown for hydrogen adsorption on Ni@Pd.

Structures and segregation patterns of Ag-Cu and Ag-Ni nanoalloys adsorbed on MgO(0 0 1).

Low-energy geometric structures and segregation patterns of Ag-Cu and Ag-Ni nanoparticles adsorbed on MgO(0 0 1) are searched for by global optimisation methods within an atomistic potential model. Sizes betwen 100 and 300 atoms are considered for several compositions. In all cases, Ag segregates to the nanoparticle surface, so that Cu@Ag and Ni@Ag core-shell arrangements are found, with off-centre cores for Ag-rich compositions. The behaviours of Ag-Cu and Ag-Ni differ at the interface with the MgO substrate. For Ag-Cu, some Cu atoms are at the interface even for compositions that are very rich in Ag, where Ag-Ni nanoparticles present an interface completely made of Ag atoms. Ag-Ni and Ag-Cu also differ concerning their geometric structures. With increasing Ag content, in Ag-Cu we find the structural sequence faulted fcc [Formula: see text] icosahedral [Formula: see text] fcc, while in Ag-Ni we find the sequence hcp [Formula: see text] faulted fcc-faulted hcp [Formula: see text] icosahedral [Formula: see text] fcc.

Calculating the free energy of transfer of small solutes into a model lipid membrane: Comparison between metadynamics and umbrella sampling.

We compare the performance of two well-established computational algorithms for the calculation of free-energy landscapes of biomolecular systems, umbrella sampling and metadynamics. We look at benchmark systems composed of polyethylene and polypropylene oligomers interacting with lipid (phosphatidylcholine) membranes, aiming at the calculation of the oligomer water-membrane free energy of transfer. We model our test systems at two different levels of description, united-atom and coarse-grained. We provide optimized parameters for the two methods at both resolutions. We devote special attention to the analysis of statistical errors in the two different methods and propose a general procedure for the error estimation in metadynamics simulations. Metadynamics and umbrella sampling yield the same estimates for the water-membrane free energy profile, but metadynamics can be more efficient, providing lower statistical uncertainties within the same simulation time.

Correction: Study of structures and thermodynamics of CuNi nanoalloys using a new DFT-fitted atomistic potential.

Correction for 'Study of structures and thermodynamics of CuNi nanoalloys using a new DFT-fitted atomistic potential' by Emanuele Panizon et al., Phys. Chem. Chem. Phys., 2015, DOI: 10.1039/c5cp00215j.

MARTINI Coarse-Grained Models of Polyethylene and Polypropylene.

The understanding of the interaction of nanoplastics with living organisms is crucial both to assess the health hazards of degraded plastics and to design functional polymer nanoparticles with biomedical applications. In this paper, we develop two coarse-grained models of everyday use polymers, polyethylene (PE) and polypropylene (PP), aimed at the study of the interaction of hydrophobic plastics with lipid membranes. The models are compatible with the popular MARTINI force field for lipids, and they are developed using both structural and thermodynamic properties as targets in the parametrization. The models are then validated by showing their reliability at reproducing structural properties of the polymers, both linear and branched, in dilute conditions, in the melt, and in a PE-PP blend. PE and PP radius of gyration is correctly reproduced in all conditions, while PE-PP interactions in the blend are slightly overestimated. Partitioning of PP and PE oligomers in phosphatidylcholine membranes as obtained at CG level reproduces well atomistic data.

Study of structures and thermodynamics of CuNi nanoalloys using a new DFT-fitted atomistic potential.

Shape, stability and chemical ordering patterns of CuNi nanoalloys are studied as a function of size, composition and temperature. A new parametrization of an atomistic potential for CuNi is developed on the basis of ab initio calculations. The potential is validated against experimental bulk properties, and ab initio results for nanoalloys of sizes up to 147 atoms and for surface alloys. The potential is used to determine the chemical ordering patterns of nanoparticles with diameters of up to 3 nm and different structural motifs (decahedra, truncated octahedra and icosahedra), both in the ground state and in a wide range of temperatures. The results show that the two elements do not intermix in the ground state, but there is a disordering towards solid-solution patterns in the core starting from room temperature. This order-disorder transition presents different characteristics in the icosahedral, decahedral and fcc nanoalloys.

Chemical ordering in magic-size Ag-Pd nanoparticles.

Chemical ordering in magic-size Ag-Pd nanoalloys is studied by means of global optimization searches within an atomistic potential developed on the basis of density functional theory calculations. Ag-rich, intermediate and Pd-rich compositions are considered for fcc truncated octahedral, icosahedral and decahedral geometric structures. Besides a surface enrichment in Ag, we find a significant subsurface enrichment in Pd, which persists to quite high temperatures as verified by Monte Carlo simulations. This subsurface Pd enrichment is stronger in nanoparticles than in bulk systems and is rationalized in terms of the energetics of the inclusion of a single Pd impurity in an Ag host nanoparticle. Our results can be relevant to the understanding of the catalytic activity of Ag-Pd nanoparticles in those reactions in which subsurface sites play a role.